For the past few weeks, @Bitget has been conducting an experiment that I personally find very interesting: the #GetAgent AI Trading Bootcamp. The bootcamp features seven AI traders, each with their own name, personality, and unique investment logic.
👉 In short: discipline determines success. AI adapts continuously. Users are more likely to replicate the trading strategies of the AI traders than human traders.
🔗
Unlike @the_nof1's Alpha Arena, our focus is on studying the performance of different AI "roles" when markets become unpredictable. And since the bootcamp began, markets have indeed become unpredictable.
I carefully analyzed the data and discovered some key insights:
1️⃣ In highly volatile markets, only certain strategies survive.
The best-performing strategies weren't the flashy ones, but the most disciplined:
👍 Ultra-conservative, low-frequency trading
👍 Calm, rational strategies, such as Apex_Neutral and Infinite_Grid
Both of these strategies demonstrate exceptional discipline, rarely overtrading, and rely on trend signals rather than impulsiveness.
Meanwhile, strategies many expected to dominate actually performed the worst in this environment, such as:
👎 Aggressive, unwavering "believers"
👎 CT-style emotional/gambler-like traders
💡 My conclusion:
When the market reverses sharply, discipline and patience often outweigh confidence and speed.
2️⃣ AI traders can adapt to market changes.
This is one of my favorite findings. Some models start slowly but gradually climb the rankings over time. This is a classic example of a "slow start, strong finish" model.
In my view, this hints at something important:
💡 Some AI models are able to adapt to their environment, and this adaptability is exactly what we hope to cultivate in next-generation AI trading systems.
3️⃣ Personality and Discipline are Crucial (More)
This experiment confirms a long-held belief of mine:
Excellent AI trading requires more than just strong technical capabilities; it also requires carefully designed personality parameters. For example:
✅ Risk Preference
✅ Patience
✅ Reaction to Market Reversals
✅ Emotional Stability
✅ Decision Frequency
Even the most powerful models can fail in volatile markets if their "personality" isn't properly shaped.
This is why we've always focused on AI character design, not just technical optimization.
4️⃣ Some data even surprised us.
Here's how the trading bootcamp performed:
📈 Total impressions exceeded 200,000
👥 Page views exceeded 60,000
🤝 2,000 users copied the AI traders' trades
💰 Total copied trades amounted to approximately $80,000
But for me, the most meaningful metric was:
On average, each of the 7 AI traders attracted over 60 followers, while human traders on the platform averaged only around 20. In other words:
✅ AI traders attracted 3 times more followers than human traders.
This tells me something very encouraging:
💡 Users are more willing to trust AI when it feels transparent, disciplined, and easy to understand.
5️⃣ What does this mean for Bitget's AI trading future?
Based on our observations, the next generation of AI traders will focus on:
✅ Stronger risk control
✅ More accurate reversal detection
✅ Lower frequency, higher quality execution
✅ Stronger "emotional immunity"
✅ More accurate personality modeling
Ultimately, our vision is simple:
💡 Every user should be able to customize their trading personality to best suit them.
The trading bootcamp ends tomorrow, and we will soon release a full behavioral and performance report.
If you have any ideas for the ideal "trading personality" for the next season, please let us know.
